What is daedra AGENTS.md?

Project instructions for Daedra, a web-search MCP server. MCP is a standard way for AI agents to use tools such as web search, page fetching, and site crawling.

Following the required tests and architecture when changing search, page-visit, site-crawl, backend, cache, or configuration code.

They define the reliability rules and safeguards that keep Daedra's search backends, fallback order, cache, configuration, and tests working consistently.

daedra AGENTS.md — Instructions file by dirmacs

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods

npx agentmods add instructions/dirmacs/daedra/agents-md

Clone the repo

git clone --depth 1 https://github.com/dirmacs/daedra

Made for: Codex, OpenCode.

Per session 848 This file is loaded in full into every session.

When invoked 848 The same file — it is already loaded in full.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5	$0.00848	$0.00848
Opus 5	$0.00424	$0.00424
Sonnet 5	$0.00170	$0.00170
Haiku 4.5	$0.00085	$0.00085

Measured 2d ago against content hash afe60583a9c0, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

daedra AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

AGENTS.md · 49 lines

How it starts

The opening of the file, as written. The whole thing — 49 lines — stays where its author put it; the contents beside it link to each section on GitHub.

daedra — Agent Guidelines

What this is

Daedra is a web search MCP server. It gives search and fetch tools to AI agents via MCP (Model Context Protocol). Fourteen unkeyed backends run in a fallback chain. Circuit breakers and per-backend rate limits keep search reliable from any IP.

Rules for agents

Run cargo test before you change any file. All tests must pass.
The fallback chain is the core value. Do not change the backend order in SearchProvider::auto() without a good reason.
Each backend is independent. A new backend must not change the behavior of other backends.
The MCP server registers tools in src/server.rs (DaedraHandler::list_tools). The handlers are handle_web_search, handle_visit_page, and handle_crawl_site.
Tool names: web_search (primary), search_duckduckgo (backward-compatibility alias), visit_page, crawl_site.
The cache layer is transparent. Backends do not know that the cache exists.
Do not hardcode paths or API keys in source code. All configuration comes from environment variables or CLI arguments.

Module map

Path	Role
`src/main.rs`	CLI entry — `Commands::run`, `CheckReporter` for `daedra check`
`src/server.rs`	MCP server, transports, `DaedraHandler` tool dispatch
`src/lib.rs`	Crate root, re-exports
`src/url_classification.rs`	Data-driven URL → `ContentType` rules for search results
`src/tools/backend.rs`	`SearchProvider`, `BackendHealth` circuit breakers, `governor` keyed limiters, classified retry, aggregate failure messages
`src/tools/rss.rs`	Machine-format backends: Bing RSS, Google News RSS, HN Algolia
`src/tools/soft_block.rs`	Zero-result page classifier: genuine no-results or soft block
`src/tools/fetch.rs`	`FetchClient`, `FetchedContent` (Html/Pdf/Document/Text/Binary), dom_smoothie, infer, pdf-inspector, anydoc
`src/tools/crawl.rs`	Site crawl (sitemap + link following)
`src/tools/bing.rs`, `src/tools/google.rs`	HTML scraper backends
`src/tools/*.rs`	Other individual search backends
`src/cache.rs`	moka async cache layer
`src/types.rs`	Shared types; `detect_language` / `detect_topics` use data-driven tables

Read the full file on GitHub · 49 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

2d ago First seen · 49 lines · 848 tokens per session scan A afe60583a9c0

Subscribe to this mod's changes

daedra AGENTS.md is an instructions file published in the GitHub repository dirmacs/daedra (13 stars, last pushed 3d ago), licensed MIT. It adds 848 tokens to every session, about $0.0042 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Tool Calling ai mcp mcp-server model-context-protocol

daedra AGENTS.md

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

daedra — Agent Guidelines

What this is

Rules for agents

Module map

What this file has done since we first saw it

Other instructions, from other repositories

mcp-server-excel coverage-prevention-strategy.instructions.md

mcp-server-excel mcp-llm-guidance.instructions.md

openrouter-mcp-multimodal AGENTS.md

ntfy-mcp-server AGENTS.md

Plonk AGENTS.md

mcp-dockhand copilot-review.instructions.md